Comparative Study on Sustainability Balanced Score Card (Susbsc) and Firm Value

The results expressed in Table 4 helps to provide some insight into the nature of the selected quoted manufacturing firms in South Africa used in this study. First, it can be observed that on the average, in a 10-year period (2009 -2018), the sampled firms in South Africa were characterized by positive TOBIN q 1.772657. This is an indication that most quoted firms in South Africa have a positive firm value. Also, the large difference between the minimum and maximum values of the total assets (FSIZE) showed that the sampled quoted firms in this study are not mainly dominated by either large or small firms and are widely dispersed. This is confirmed by the wide variations recorded in the standard deviation values of the variables used.

Furthermore, the Table 4 shows that on the average of ten years period, that the firms sampled are characterized by positive Balanced scorecard financial perspective value of (ROA=6.652286). This shows that during the period under study, the financial measures that convey the economic consequences of the action undertaken by the sampled firms that focused on profitability options were positively recorded. However, the wide variation between the maximum and minimum values of balanced scorecard financial perspective (ROA) which stood at 36.24000 and -38.10000 respectively justifies the need for this study as we expect that those firms with higher balanced scorecard financial perspective will have higher firm value than those with smaller or negative balanced scorecard financial perspective value.

Similarly, Balanced Scorecard Internal Business Perspective (INTBP) value on the average stood at 36.99643. This shows that large number of our sampled firms recorded positive INTBP value during the period under study. In other words, large numbers of the sampled firms were engaged with activities that can result into financial success and customer‟s satisfaction, such as cost and quality assurance activities, relating to the business process. Although, the maximum and minimum values of INTBP shows a wide variation as it stood at 95.38000 and 2.480000 respectively. This wide variation also justifies the need for this study as we assume that firms with higher balanced scorecard internal business perspective will have higher firm value than those with less Internal Business Perspective (INTBP).

Furthermore, Balanced Scorecard Customers Perspective (CUSP) shows a positive value of 9.865771. This means that most of our sampled firms were involved in activities that provide quality goods and services, ensure effective delivery of the goods and services and ensures that customers satisfaction are sustained. However, there is a high variation between the maximum value of CUSP that stood at149.8700 and minimum value that stood at-97.60000. This wide variation in CUSP values among the sampled firms justifies the need for this study as we assume that firms with higher CUSP value will have higher firm value than those firms with low CUSP value. The same assumption is also expected from the value of Learning and Development Perspective (LDP) and balanced scorecard Sustainability Perspective (SUSBSC) respectively as they both show wide variations as well in their maximum and minimum values respectively.

Lastly, in table 4, the Jarque-Bera (JB) which test for normality or the existence of outliers or extreme values among the variables shows that all the variables are distributed normally at the 1% level of significance except Balanced Scorecard Internal Business Perspective (INTBP) which is normally distributed at 10% level of significance. This implies that any variable with outlier are not likely to distort our conclusion and are therefore reliable for drawing generalization. This also implies that the least square estimation can be used to estimate the pooled regression model. The study on trying to diagnose for the presence of multicolinearity in our data used, as well as evaluating the association among the variables adopted, employed the Pearson correlation coefficient (correlation matrix) analysis. The result obtained is presented in Table 5

Source: Researchers Computation (2019)

The use of correlation matrix in most regression analysis is to check for multi-colinearity and to explore the association between each explanatory variable (ROA, INTBP, CUSP, LDP, SUSBSC, and FSIZE) and the dependent variable (TOBINQ). Table 5 focused on the correlation between firm value measured as TOBINQ and the independent variables (ROA, INTBP, CUSP, LDP, SUS, and FSIZE) of the quoted manufacturing firms in South Africa. The finding from the correlation matrix table shows that all our independent variables, (ROA = 0.17; INTBP = 0.04; CUSP=0.03, LDP=0.06, SUS=0.03 and FSIZE=0.10) were observed to be positively and weakly associated with Firm value measured as TOBIN q. In checking for multi-collinearity, we notice that no two explanatory variables were perfectly correlated. This means that there is no problem of multi-collinearity between the explanatory variables. Multi-collinearity usually results to wrong signs or implausible magnitudes in the estimated model coefficients obtained. There will also be bias in the standard errors of the coefficients.

MODEL 2: General Model Specified for South African Listed Manufacturing Firms:

In other to examine the impact relationships between the dependent variable TOBINQ and the independent variables (ROA, INTBP, CUSP, LDP, SUSBSC, and FSIZE) and to also test the formulated hypothesis given, the study used a pooled multiple regression analysis, to test the hypothesis, owing to the fact that the data had both time series (20092018) and cross sectional properties (35 quoted manufacturing firms in South Africa). The first pooled interaction based multiple regression result we obtained shows a result that has an autocorrelation problem, just exactly what we observed from data gathered from quoted manufacturing firms in Nigeria.

Auto correlation problem is a problem that is predominantly with time series data and our data also have a time series characteristics as it cut across ten (10) years period, 2009-2018. The autocorrelation problem in our regression result was confirmed by the Durbin-Watson value of the result being 0.435336 which is less than 2 or approximately 2 that is the rule of thumb

In other to correct for the autocorrelation problem in the result, we can use different methods such as using CruchGodfrey L-M test, or re-estimated the regression equation by introducing AR variable into the estimate. But for the purpose of this study, we corrected the autocorrelation problem observed in our regression result by re-estimating our equation with AR variable into the estimate and the Generalised Least Square (GLS) result obtained as a result of the correction is presented as table 6 and is interpreted below

Table 6: TOBIN q Pooled Generalized Least Square (GLS) Regression Result From Data Collected from Quoted Manufacturing Firms in South Africa.

Cross-sections included: 35

Total panel (balanced) observations: 315

Convergence achieved after 6 iterations

Comparative Study on Sustainability Balanced Score Card (Susbsc) and Firm Value

Source: Source: Researchers’ summary of South Africa firms analysis (2019) from E-view 9.0 statistical package

In table 6, R-squared and its adjusted R-squared values were (0.89) and (0.89) respectively. This is an indication that all the independent variables jointly explain about 89% of the systematic variations in Firm value (TOBIN q) of our sampled companies over the ten-year period (2009-2018) while 11% of the systematic variations are captured by the error term. The F-statistics 361.5890 and its P-value of (0.00) portrays the fact that the TOBIN q regression model is well specified.

Test of Autocorrelation: Using Durbin Watson (DW) statistics which we obtained from our regression result in table 6, it is observed that DW statistic is2.375584 which is approximately 2, agrees with the Durbin Watson rule of thumb. Showing that our data are free from autocorrelation problem and as such fit for the regression result to be interpreted and result relied on. Akika Info Criterion and Schwarz Criterion which are 2.110399 and 2.205702 respectively further strengthen the fitness of our regression result for reliability as they confirm the goodness of fit of the model specified. In addition to the above, the specific findings from each explanatory variable are provided as follows:

Balanced Scorecard Sustainability Perspective (SUS), based on the t-value of -0.045793 and P-value of 0.96 in table 6 above, was found to have a negative influence on our sampled quoted manufacturing firm‟s value (TOBIN Q) in South Africa and this influence is statistically insignificant since its P-value is more than 10%. This result, therefore suggests that we should accept our null hypothesis (H0) which states that Sustainability Balanced Scorecard (SUS) does not have significant relationship with firm value. This means that quoted manufacturing firms in South Africa that engages more on environmental sustainable activities are rated lower in scorecard value by their investors and shareholders and this higher sustainability activity therefore drives the value of their firms negatively. However, this influence is not statistically significant and therefore, should be ignored by managements. This result by implication shows that investors and shareholders in South Africa does not want the managers to excessively engage on activities that will not bring immediate gain to them or increase their dividend base. Therefore, as this influence is not statistically significant, management should pay more attention on the activities that will increase firms profitability and less attention on environmental sustainability activities but since sustainability activities does not significantly influence firm value in South Africa, management as well as policy makers should ignore it, when considering activities that can drive balanced firm values among manufacturing firms in South Africa.

Test of Control variable: Firm size (FSIZE) and Firm Value (TOBIN q), based on the t-value of 0.091169 and P-value of 0.93 in Table 6 above, was found to have a positive influence on the quoted manufacturing firm‟s value (TOBIN q) in South Africa and this influence is statistically insignificant because the p-value is more than 10%. This invariably means that in terms of rating of firm‟s value by investors and shareholders, firms with large sizes are rated higher than firms with small sizes. Therefore, as this influence is statistically insignificant, management should pay less attention on the activities that will increase their sizes more such as investment and expansion activities, since it does not significantly affect the value of their firms in South Africa. This means that large firm sizes (FSIZE) does not drives the value of firms significantly and could be ignored by the managements.

Comparative Analysis of Countries Specific Results

The result provided an insight into the nexus between balanced scorecard and firm values of manufacturing firms quoted across countries used for this study (Nigeria and South Africa) and the result is presented in table 7